(1) Field of the Invention
The present invention generally relates to vertically polarized RF antennas and more specifically to bicone and dipole antennas.
(2) Description of the Prior Art
In the signal detection environment, it is often desirable to detect both horizontal and vertical polarized signals on the horizon over a broad bandwidth. It is also desirable that the response to both polarizations is equal. Optionally, an antenna may be used in a vertical stack of antennas, which requires cables to pass vertically by the antenna to other antennas above the antenna.
A typical solution to detecting both horizontal and vertical polarized signals over a broad bandwidth whereby the response to both polarizations is equal, is to use a vertically polarized bicone antenna positioned cylindrically in many cylindrical layers of polarizing grids that slowly rotate the incident field from cross 45 degrees to close to cross 0 degrees (90 degrees to the horizon, vertically polarized), as shown in FIGS. 1A and 1B. Grids pass signals that are cross or perpendicular to, the grid angle or pitch angle of the grids.
The components of to-be-received horizontal and vertical polarized signals of equal amplitude are of equal amplitude at cross 45 degrees. Thus for equal response to either polarization, the grid and bicone configuration receives this component of either of the polarizations. A bare minimum of three layers of grids of grid angles 45, 30 and 15 degrees rotates the cross 45 degree component from respectively cross 45 to cross 30, and to cross 15.
The bicone inside of the 15 degree grid layer finally receives the signal at cross 0 degrees (vertical polarization, 90 degrees to the horizon). One layer of 45 degree grids is a possible configuration that passes a to-be-received signal component of cross 45 degrees to the cross 0 degrees vertical polarization of the bicone, although there is a 3 decibel polarization mismatch loss.
In general, the amount of rotation of polarization to be done by the grids determines the bare minimum number of grid layers that is needed to minimize to a reasonably small value the polarization mismatch loss of the grids. A reasonable rule dictates that at least one grid layer is needed per 15 degrees of rotation. Thus 45 degrees of rotation requires three grid layers with grid angles of 45, 30 and 15 degrees.
The grid layers are arranged radially so that the layers of larger grid angles have corresponding larger radii. In other words, a grid layer's pitch angle will increase directly with an increase in radius.
In all configurations, with the length of the grid lines long enough to cause reflections, the distance between adjacent grid layers must be less than ½ wavelength (½λ) and the distance between the innermost grid layer and the axis of the bicone must be less than 1 wavelength. At ½ or 1 wavelength separations, cancellation of the incident and reflected signals occurs respectively between adjacent grid layers or the innermost grid layer and the axis of the bicone, causing a problematic signal gain null and degraded patterns. For the innermost grid layer and the axis of the bicone case, this null also occurs at higher integer multiples of a half wavelength, e.g., 1.5, 2.5 . . . wavelengths, although of decreasing intensity as the integer multiples of a half wavelength increases.
NEC (Numeric Electromagnetic Code) government models using a narrow dipole antenna instead of a bicone antenna in the configuration have shown that cancellation between the innermost grid layer and a dipole starts at 0.5, not 1.0, wavelength separation. It is not known if the difference between the two cases is dependent upon the use of a bicone or a dipole.
Resultant problems and additional problems of using grids external to the bicone are noted below.
The distance between the innermost grid and a bicone axis can quickly approach 1 wavelength at which cancellation occurs, especially if the bicone is being used at frequencies where it is electrically large, i.e., larger than 1 wavelength.
Additionally, the grids are parasites of finite size and thus can have resonances, e.g., circumferential resonances in this configuration.
Circumferential resonances occur when the circumference of the grid layer is an integer multiple of a wavelength. Such resonances usually severely degrade the required azimuthal omni patterns of the bicone at the frequencies of the resonances, due to re-radiation or reflection off of the grids. The resonances appear as negative spikes or nulls when illustrated in gain versus frequency plots.
Resonances can be expected to be made less severe by making the grids more broadband by making them larger. For example, the heights of the grids can be increased. However, measurements have indicated that increasing grid height has little effect on decreasing pattern degradation at resonance frequencies.
Additionally, the circumference of grid layers can be made larger so that higher order resonances appear in the frequency band of interest. The worse case resonance is when the circumference is 1 wavelength. Usually resonances when the circumference is 3 or more wavelengths can be ignored since pattern degradation is small.
Typical measurements, such as for a configuration of a bicone inside 15 and 30 degree layers of grids, have shown the following severity of gain nulls (see TABLE 1).
TABLE 1ResonanceCircumference ofOutermost Grid LayerGain Null Depth(wavelengths)(decibels)1−352−7
Vertical and horizontal polarization gain (decibel) versus frequency (wavelength) plots for all azimuths are shown in respective FIGS. 2 and 3. First and second nulls are noted for each grid layer, N115, N130, N215 and N230, respectively. The frequencies in FIG. 2 and FIG. 3 corresponding to the respective 1 and 2 wavelength circumferences of the 15 degree grid layer are noted at 1λ15° and 2λ15° whereas the frequencies corresponding to the respective 1 and 2 wavelength circumferences of the 30 degree grid layer are noted at 1λ30° and 2λ30°. Notice that nulls are located at or close to the circumferential wavelength points.
Note that a small space separates the 15 degree and 30 degree grids with the 30 degree grids being a little larger in radius and circumference. This results in the resonances of both grids not being exactly at the same frequency. The 30 degree grid resonances are a little lower in frequency than those of the 15 degree grids. The 1λ and 2λ resonance frequencies for both grids are shown on FIGS. 1 and 2. The double resonances are seen with the first null being actually composed of two closely spaced nulls N115 and N130, and the second null composed of two closely spaced nulls N215 and N230.
The nulls are at or close to the circumferential wavelength points. Gains for horizontal polarization and below the first null are spread out since the antenna configuration was below its cut-in frequency resulting in the feed cable of the bicone starting to become part of the antenna and radiating asymmetrically in the azimuth plane. Measurement problems also spread out the gains above the second null.
Optionally, cables are allowed to vertically pass the antenna configuration. To allow one or more cables to pass vertically by the antenna, the cables are run parallel to the outermost layer of grids and outside of, and insulated from, the grids as shown in FIG. 1A.
Running the cables parallel to the outermost layer of grids minimizes shunting of the antenna by the cables and upsetting the azimuth patterns, since the cables are at the polarization shunted by the grids. To minimize further shunting by the cables in the area past the grids, the cables should be kept away from the top and bottom of the bicone, and thus should maintain at least the same radial distance from the bicone axis as in the area where the cables pass the outermost grid layer.
Describing the components of the antenna configuration in FIG. 1A in more detail, a bicone antenna 30 is made up of two cones, a top cone 18 and a bottom cone 20 arranged as shown. A coaxial cable 22 feeds the bicone at the feed point 26 of the antenna where the outer conductor 21 of the cable 22 is connected to the bottom cone 20 and the center conductor 23 of the cable 22 is connected to the top cone 18. The feed angle θF is the angle between the two cones at a feed point 26, and is usually picked so that the characteristic impedance Z0 of the bicone is that of the coaxial cable (50 ohms) to ensure optimal match.
The bicone is wrapped in three layers 12, 14, 16 of grids of grid angles of 15, 30 and 45 degrees (see also FIGS. 1B, 1C, 1D), respectively, where the separation between adjacent grids or the innermost grid and the extreme edge of the bicone is a constant value. The grid layers are arranged radially so the layers of larger grid angles are of larger radius from the antenna axis 10.
The grid layers are kept separated by placing foam spacers between adjacent layers or between the innermost layer and the bicone, or by placing the whole configuration in a cylindrical dielectric box (not shown) whose top and bottom are mounted to the top and bottom of the bicone and whose top and bottom have slits on their insides in which the top and bottom edges of the grid layers are placed and held in place.
The height 24 of the grid layers are at least that of the height 24 of the bicone (FIG. 1A). Any cables 28 running vertically past the antenna 30 are run parallel to the grids of the outermost layer of grids and outside of, and insulated from the grids. In the same area past the grids, the cables preferably maintain at least the same radial distance from the bicone axis 10 as in the area where the cables pass the outermost grid layer. If there is more than one cable, the cables are placed symmetrically about the circumference of the outermost grid layer.
FIGS. 1B, 1C and 1D details the layers 12, 14, 16 of FIG. 1A with each layer unwrapped. A layer is made by etching or cutting metal 41 off of a plastic sheet 43. A typical way to un-join the layer from its cylindrical position about the bicone is to cut the layer along a line 40 parallel to and halfway between two of the grids making up the layer. It can be rejoined later with tape.
A grid 45 is a metal line that extends from the bottom to the top of the sheet having a height 24. The angle the grid makes with the horizontal is the grid angle θG. In FIGS. 1B, 1C, 1D the grid angle θG is 45, 30 and 15 degrees, respectively. The width 32 of the grid compared to the width 34 of the non-gridded area of the sheet is somewhat arbitrary. The width 32 is non-critical as long as the extremes of the line width are not used. In FIG. 1B the metal width divided by the total width 36 available for a grid is 0.5.
The length of a grid line should be at least ½ wavelength, so it can reflect away the field parallel to the grids.
The number of grids on a sheet is determined by the required spacing between the grids. This spacing should be appreciably less than 1 wavelength for appreciable attenuation of the field parallel to the grids, as further discussed in the case of a bicone in internal grids below. Note that the actual number of grid lines is much larger than shown in FIGS. 1B to 1D.
Another possibility to reduce the effects of possible resonances and to reduce pattern degradation when the separation between the bicone and grids is one wavelength is to place the grids within the bicone as shown in FIG. 4.
The internal grid bicone antenna 51 of FIG. 4 includes a bicone having a height 59, two circumferential edges 57, and a radius 15 measured from the bicone or antenna axis 10. The antenna includes a top cone 18, a bottom cone 20, a feed angle θF, a feed point 26, 15 degree grid layer 50, 30 degree grid layer 52 and 45 degree grid layer 54.
An internal grid bicone antenna reduces resonance effects since the grids are more highly coupled to the bicone and are thus less a parasite. If the layers are small enough in circumference, operation below the first resonance, where the circumference is one wavelength, is possible. Also the null frequency, where the separation between the bicone axis and innermost grid is one wavelength, is pushed to higher frequencies since the separation is reduced. Thus the antenna configuration can be used at higher frequencies. However, several problems arise from this configuration.
Placing the grids within the bicone starts to defeat the purpose of the grids, which is to rotate the field for an antenna behind the grids. In the internal grid configuration, part of the bicone is behind the grids in the rotated field, and part of the bicone is ahead of the grids in the field that has not yet been rotated.
The part of the bicone ahead of the grids is the two horizontal edges 57 of the bicone, which act as a pair of horizontal grids. Long grids start to allow the field they are supposed to block, i.e., the field parallel to the grids, to pass at roughly when the spacing between the parallel grids is 1 wavelength. Thus, below 1 wavelength the two horizontal edges 57 of the bicone can block horizontal polarization from being received by the grids and bicone. Thus the height 59 of a wide internal grid bicone must be at least 1 wavelength for horizontal response to equal vertical response.
For an idea of the approximate amount of blockage that a layer of grids can achieve to polarization parallel to the grids, below is a representative TABLE 2 of the amount of field blockage of grids versus grid separation obtained from NEC from a dipole probe in front of a 2 wavelength by 2 wavelength sheet of grids.
TABLE 2Grid SeparationNumber of Grids(wavelengths)Rejection (db)310.5(interpolated)51/2263/82.591/44.5171/89.533 1/1616
The cases with a small number of grids in Table 2 above may be suspect since only a 2 wavelength by 2 wavelength sheet of grids was used.
Placing the grids within the bicone increases the cut-in frequency of the bicone more than what occurs with bicones with external grids. The cut-in frequency is defined as the frequency where the VSWR (Voltage Standing Wave Ratio) about the characteristic impedance Zo of the antenna becomes a low, flat level. This is because the grids are shunting the bicone at a point closer to the bicone feed point 26.
Vertical passage of cables past a bicone in either external or internal grids is done by having the cable follow the path of the grids of the outermost grid layer outside of the outermost grid layer. Since grids shunt any field along their direction, the antenna impedance-wise will usually not see the cable at the frequencies where it is of low impedance, being those frequencies above cut-in. Below cut-in, the bicone and similarly sized grids have high impedance and thus coupling to lower impedance longer cables going past the bicone is possible resulting in the cables becoming part of the antenna and radiating usually undesirable patterns. Obviously, this is an undesirable property if operation below cut-in is important.